Sikorsky X2 coaxial heli developments.
The Sikorsky X2 TECHNOLOGY™ demonstrator aircraft will incorporate several new technologies and demonstrate them in a flight environment. These technologies include an integrated Fly-by-Wire system that allows the engine/rotor/propulsor system to operate efficiently, with full control of rotor rpm throughout the flight envelope, high lift-to-drag rigid blades, low drag hub fairings, and Active Vibration Control.
-- IFMU
Join Date: Nov 2004
Location: Cambridgeshire, UK
Posts: 1,334
Likes: 0
Received 0 Likes
on
0 Posts
Helicopter vibration basics for the not of faint heart.
Nice to see that the X2 discussion is still along the lines of technical merit of X2. 
Nick, your occasional considered comments are very welcome. Hopefully the thread will stay non-political so that any technical feedback you care to offer will not tread on the toes of other well considered high speed helicopter development.
----
An explanation of just how complicated helicopter vibration can become!
The best analogy of how a blade behaves in flight is to look at a skipping rope dangling in the effect of gravity (ie one end free the other held). If you wiggle the end at various speeds you can induce any number of tension wave modes along the fixed length. In a rotor this tension is clearly generated by the centrifugal force. The tension waves are also being sped up by bending waves which are themselves a combination of shear waves (wavespeed independent of frequency) and flexural waves (wavespeed proportional to frequency^0.5), but this only really alters the resonance frequency at which each tension wave occurs.
The fuselage will potentially respond to any of the the blade modes transmitted through the hub. Ideally the number of blades and Nr is chosen to avoid any excitation of the structural modes of the fuselage. The combination of blades resonating around the azimuth at relative phases can produce symmetric modes and antisymmetric modes in the axis symmetry of the rotorshaft. The fuselage sees the symmetric modes acting vertically on the rotor head, but for antisymmetric modes there is the complication that the rotor is in a rotating frame while the fuselage is in a static frame. This means that the antisymmetric mode frequency will apply a bending moment to the rotorhead with two simultaneous frequencies, shifted higher and lower by Nr/60 Hz than the original resonant frequency in the blades. To explain this would require use of DeMoivres elegant theorem, multiplying two Euler number representations of sinusoids for rotating and static frame. Anyone up for imaginary numbers?
Of course all of this is then actually excited by the cyclically varying aerodynamic loads, which will increase as the helicopter speeds up and the swash plate moves to keep the blades flapping to equality...
----
Suffice to say the X2 rotor dynamicists will have been extremely busy to get all of this right!
On X2 the vibration is absorbed at (or close to) the hub by phased eccentric balancers. This means that even if there was a potential to excite the fuselage then the Moog system should take care of it. The best measure of blade dynamics is really how the tips move, for which there are sensors in X2. Even the tail prop behaviour in the pulsing rotor downwash is likely to have been studied in some detail in either testing of coupled CFD-FEA analyses. So the flight test program is ideally just looking for unexpected interactions.
Hopefully, all will go smoothly when new gearbox is fitted up. But don't expect anything to happen quickly. The reason for flight test is to explore the unknown, so the best way to proceed is very carefully. That way by the time this technology becomes commercially available the major risks will have been found and corrected. This is true of any well managed project...
Nick, your occasional considered comments are very welcome. Hopefully the thread will stay non-political so that any technical feedback you care to offer will not tread on the toes of other well considered high speed helicopter development.
----
An explanation of just how complicated helicopter vibration can become!
The best analogy of how a blade behaves in flight is to look at a skipping rope dangling in the effect of gravity (ie one end free the other held). If you wiggle the end at various speeds you can induce any number of tension wave modes along the fixed length. In a rotor this tension is clearly generated by the centrifugal force. The tension waves are also being sped up by bending waves which are themselves a combination of shear waves (wavespeed independent of frequency) and flexural waves (wavespeed proportional to frequency^0.5), but this only really alters the resonance frequency at which each tension wave occurs.
The fuselage will potentially respond to any of the the blade modes transmitted through the hub. Ideally the number of blades and Nr is chosen to avoid any excitation of the structural modes of the fuselage. The combination of blades resonating around the azimuth at relative phases can produce symmetric modes and antisymmetric modes in the axis symmetry of the rotorshaft. The fuselage sees the symmetric modes acting vertically on the rotor head, but for antisymmetric modes there is the complication that the rotor is in a rotating frame while the fuselage is in a static frame. This means that the antisymmetric mode frequency will apply a bending moment to the rotorhead with two simultaneous frequencies, shifted higher and lower by Nr/60 Hz than the original resonant frequency in the blades. To explain this would require use of DeMoivres elegant theorem, multiplying two Euler number representations of sinusoids for rotating and static frame. Anyone up for imaginary numbers?
Of course all of this is then actually excited by the cyclically varying aerodynamic loads, which will increase as the helicopter speeds up and the swash plate moves to keep the blades flapping to equality...
----
Suffice to say the X2 rotor dynamicists will have been extremely busy to get all of this right!
On X2 the vibration is absorbed at (or close to) the hub by phased eccentric balancers. This means that even if there was a potential to excite the fuselage then the Moog system should take care of it. The best measure of blade dynamics is really how the tips move, for which there are sensors in X2. Even the tail prop behaviour in the pulsing rotor downwash is likely to have been studied in some detail in either testing of coupled CFD-FEA analyses. So the flight test program is ideally just looking for unexpected interactions.
Hopefully, all will go smoothly when new gearbox is fitted up. But don't expect anything to happen quickly. The reason for flight test is to explore the unknown, so the best way to proceed is very carefully. That way by the time this technology becomes commercially available the major risks will have been found and corrected. This is true of any well managed project...
Last edited by Graviman; 12th Mar 2010 at 09:47. Reason: To make content readable, but this just is a complicated subject.
XH-59 Vibration Absorber System
Dave Jackson,
The following note from Andy Ruddell, who ran the aero and rotors side of that program, provides the story:
You brought back a flood of memories John. Yes there was one designed (and built). It happened during the great rotor index debate - 0 deg vs 60 deg. At 60 deg the predominant vibration is lateral and longitudinal at 0 deg. The aircraft was phased at 0 deg when you flew it. A lateral absorber was designed, built and shipped to West Palm. It was big enough to almost fill the entire cabin. It was never installed. Shortly thereafter we went to 0 deg phase. That is a whole other story.
Thanks,
John Dixson
The following note from Andy Ruddell, who ran the aero and rotors side of that program, provides the story:
You brought back a flood of memories John. Yes there was one designed (and built). It happened during the great rotor index debate - 0 deg vs 60 deg. At 60 deg the predominant vibration is lateral and longitudinal at 0 deg. The aircraft was phased at 0 deg when you flew it. A lateral absorber was designed, built and shipped to West Palm. It was big enough to almost fill the entire cabin. It was never installed. Shortly thereafter we went to 0 deg phase. That is a whole other story.
Thanks,
John Dixson
Join Date: Apr 2003
Location: Vancouver, BC, Canada
Posts: 1,635
Likes: 0
Received 0 Likes
on
0 Posts
John, thanks for the information.
Mr Ruddell wrote a number of extremely interesting technical documents on the craft back then. In fact, his documents significantly motivated my interest in the ABC concept.
Dave
Mr Ruddell wrote a number of extremely interesting technical documents on the craft back then. In fact, his documents significantly motivated my interest in the ABC concept.
Dave
Join Date: Sep 2007
Location: USA
Posts: 601
Likes: 0
Received 0 Likes
on
0 Posts
Other sources of vibration
The US Army AATD is also getting serious about sources of NVH other than the rotor system. They just issued a $20 million RFP for build and test of drivetrain hardware that (among other things) can demonstrate an 18dB reduction in drivetrain generated noise versus a baseline year 2000 Army rotorcraft. That's pretty significant. Especially since most MRGB's are hard mounted to the airframe directly above the crew area, and thus efficiently transmit annoying, high-frequency drivetrain disturbances through the structure.
Should be good for Army aircrews and would definitely have commercial spin-offs.
Should be good for Army aircrews and would definitely have commercial spin-offs.
AATD Drivetrain RFP
The proposal asks for a 55% increase in the drivetrain hosepower to weight ratio compared to US Army year 2000 levels. Any of the drivetrain engineers out there think this is a practical goal?
Thanks,
John Dixson
Thanks,
John Dixson
Last edited by JohnDixson; 12th Mar 2010 at 15:52. Reason: typo in baseline year
Join Date: Nov 2000
Location: Indianapolis, IN, USA
Posts: 151
Likes: 0
Received 0 Likes
on
0 Posts
Sounds like an objective reqt. RDS-21 asked for a 35% delta in power/weight, but the split torque face gear xmsn which eventually emerged from the effort for AB3 has an improvement closer to 20%.
Join Date: Apr 2003
Location: Vancouver, BC, Canada
Posts: 1,635
Likes: 0
Received 0 Likes
on
0 Posts
Riding the Clutch
CEFOSKEY
Speculation #4;
Another attempt at uncovering the source of the elusive shavings.
Sikorsky's patent US 7,296,767 is for a "Variable speed transmission for a rotary wing aircraft'. As in the previously mentioned patent US 7,651,050, this patent also states 'variable speed' 25 times.
The following sketch is from this patent.
This patent actually describes a two-speed transmission. The rotors turn at a high speed when the clutch is engaged and at a low speed when the clutch is disengaged.
The only way that these rotors can operate at speeds in-between the high and the low is by having the clutches slip.
Does 'riding the clutch' produce the so-called 'Variable Speed'?
Did 'riding the clutch' produce the shavings?
If speculation #4 is correct then just transfer this emoticon to your reply. ~
Dave
Speculation #4;
Another attempt at uncovering the source of the elusive shavings.
Sikorsky's patent US 7,296,767 is for a "Variable speed transmission for a rotary wing aircraft'. As in the previously mentioned patent US 7,651,050, this patent also states 'variable speed' 25 times.
The following sketch is from this patent.
This patent actually describes a two-speed transmission. The rotors turn at a high speed when the clutch is engaged and at a low speed when the clutch is disengaged.
The only way that these rotors can operate at speeds in-between the high and the low is by having the clutches slip.
Does 'riding the clutch' produce the so-called 'Variable Speed'?
Did 'riding the clutch' produce the shavings?
If speculation #4 is correct then just transfer this emoticon to your reply. ~
Dave
Why would you need to continuously vary the speed for a rotorcraft? That sounds more like golf cart technology to me. Perhaps two speed on that patent means two speed.
-- IFMU
-- IFMU
Join Date: Mar 2010
Location: The Northeast
Posts: 2
Likes: 0
Received 0 Likes
on
0 Posts
To build on what Graviman said and to speculate based on my knowledge of Dynamics, variable rotor speed is probably the best way to go for a high speed coaxial helicopter, but it is not without its difficulties.
If you want to run at one rotor speed for hover up to 250+ kts, the rotor speed will either be so slow that hover performance is lousy (max thrust roughly proportional to the square of RPM) or there will be extremely high drag on the advancing blades at high speed due to compressibility. Neither of these cases is satisfactory and using a high rotor speed in hover and lower in high speed flight solves both problems.
A rotor blade has natural frequencies in both bending directions (beamwise and chordwise), and torsion too. ALL of them change with rotor speed. The issue with vibration is that the blade are excited by airloads at harmonics of the rotor speed. If a rotor turns at 10 Hz (600 RPM) then the blades will be forced at 10 Hz, 20 Hz, 30Hz, 40Hz, etc. A rotor blade had better not have any natural frequencies at those multiples at its intended operating speed, especially if one wants to cut down on cabin vibration and increase blade service life. The big problem with variable rotor speed is that the blade natural frequencies have to be kept away from those multiples for ALL intended rotor speeds. Keeping them located correctly is hard enough at one rotor speed, let alone a range.
If you want to run at one rotor speed for hover up to 250+ kts, the rotor speed will either be so slow that hover performance is lousy (max thrust roughly proportional to the square of RPM) or there will be extremely high drag on the advancing blades at high speed due to compressibility. Neither of these cases is satisfactory and using a high rotor speed in hover and lower in high speed flight solves both problems.
A rotor blade has natural frequencies in both bending directions (beamwise and chordwise), and torsion too. ALL of them change with rotor speed. The issue with vibration is that the blade are excited by airloads at harmonics of the rotor speed. If a rotor turns at 10 Hz (600 RPM) then the blades will be forced at 10 Hz, 20 Hz, 30Hz, 40Hz, etc. A rotor blade had better not have any natural frequencies at those multiples at its intended operating speed, especially if one wants to cut down on cabin vibration and increase blade service life. The big problem with variable rotor speed is that the blade natural frequencies have to be kept away from those multiples for ALL intended rotor speeds. Keeping them located correctly is hard enough at one rotor speed, let alone a range.
Join Date: Apr 2003
Location: Vancouver, BC, Canada
Posts: 1,635
Likes: 0
Received 0 Likes
on
0 Posts
IFMU,
The accepted definition for a variable speed drive is "a mechanism that is used to transmit motion from one shaft to another which allows the velocity ratio of the shafts to be varied continuously".
The X2 is probably intended to cruise at any speed between hover and maximum forward velocity. Therefore, an optimal situation will be to have the propeller start turning slowly near transitional forward velocity. Then, as the craft accelerates toward maximum forward velocity, the rotational speed and the pitch of the main rotors will proportionately decrease, while the rotational speed and the pitch of the propeller proportionately increases.
This is significantly better than 'shifting gears' on a two-speed gearbox. However, a problem with variable speed power transmission devices is that they consume an unacceptable amount of the power for helicopters, IMHO.
This being the case, why does Sikorsky incorporate the phrase 'variable speed' so widely throughout the above patents? Could it be an attempt to fight off others who seek to develop true 'variable speed transmissions, ( including mine)?
Dave
Why would you need to continuously vary the speed for a rotorcraft?
The X2 is probably intended to cruise at any speed between hover and maximum forward velocity. Therefore, an optimal situation will be to have the propeller start turning slowly near transitional forward velocity. Then, as the craft accelerates toward maximum forward velocity, the rotational speed and the pitch of the main rotors will proportionately decrease, while the rotational speed and the pitch of the propeller proportionately increases.
This is significantly better than 'shifting gears' on a two-speed gearbox. However, a problem with variable speed power transmission devices is that they consume an unacceptable amount of the power for helicopters, IMHO.
Perhaps two speed on that patent means two speed.
Dave
Last edited by Dave_Jackson; 22nd Mar 2010 at 03:41. Reason: spelling
Join Date: Nov 2004
Location: Cambridgeshire, UK
Posts: 1,334
Likes: 0
Received 0 Likes
on
0 Posts
asianj0e,
Talking about torsional and bending frequencies excited by aerodynamic loads might set folks hearts a flutter . Most folks eyes seem to glaze over when the subject of order shifting for antisymmetric modes comes up...
Dave,
AsianJoe has it spot on when he talks about avoid rpms. If there was a need to have more than two rpms then the lightest way is a multispeed epicyclic transmission. In helicopters the engine speed is sometimes varied as a function of indicated airspeed.
Does X2 not have autorotatation for cruise-dash speed ranges?
This would allow any rotor rpm...
Talking about torsional and bending frequencies excited by aerodynamic loads might set folks hearts a flutter
. Most folks eyes seem to glaze over when the subject of order shifting for antisymmetric modes comes up...Dave,
AsianJoe has it spot on when he talks about avoid rpms. If there was a need to have more than two rpms then the lightest way is a multispeed epicyclic transmission. In helicopters the engine speed is sometimes varied as a function of indicated airspeed.
Does X2 not have autorotatation for cruise-dash speed ranges?
This would allow any rotor rpm...
Join Date: Sep 2007
Location: USA
Posts: 601
Likes: 0
Received 0 Likes
on
0 Posts
Sikorsky xmsn patent
Thar Sikorsky patent describes a two speed transmission. It gets variable speed the same way the V22 does, by changing engine speed. Turboshaft engines tend to have a narrow range of usable operating speeds. Usually something like 70 to 100 percent under ideal conditions. A two speed transmission would essentially double that range.
However, that transmission schematic shows two types of clutches, a friction clutch and a sprag clutch. And each of those clutches will likely have some reliability issues the way they are being employed.
The friction clutch is being used to synchronize two shafts when the high speed gear ratio is to be engaged. The friction clutch must provide slip until the two shaft speeds are equal, which requires the clutch pack to absorb a great deal of heat energy without failure. That requires a large mass of friction material like CRC. Also, being a device that transmits power solely through friction and not having a reliable way to monitor its condition, good design practice would probably dictate that the friction clutch should be mechanically shunted (with a dog clutch or some similar positive engagement device) once it has fully synchronized.
As others have noted, driveline shaft dynamics can be a real nightmare. The varying and aggressive friction characteristics of carbon materials makes it near impossible to smoothly modulate the slip of a carbon clutch pack. This issue was a real problem for Rolls/Allison on the F-35 lift fan clutch.
As for the sprag clutch, they do not like to operate under overrun for extended periods, such as this Sikorsky design concept would require.
Since they have two engines with an interconnect, they would be better off doing a fully synchronized shift using dog clutches with the speeds controlled by the FADEC's.
However, that transmission schematic shows two types of clutches, a friction clutch and a sprag clutch. And each of those clutches will likely have some reliability issues the way they are being employed.
The friction clutch is being used to synchronize two shafts when the high speed gear ratio is to be engaged. The friction clutch must provide slip until the two shaft speeds are equal, which requires the clutch pack to absorb a great deal of heat energy without failure. That requires a large mass of friction material like CRC. Also, being a device that transmits power solely through friction and not having a reliable way to monitor its condition, good design practice would probably dictate that the friction clutch should be mechanically shunted (with a dog clutch or some similar positive engagement device) once it has fully synchronized.
As others have noted, driveline shaft dynamics can be a real nightmare. The varying and aggressive friction characteristics of carbon materials makes it near impossible to smoothly modulate the slip of a carbon clutch pack. This issue was a real problem for Rolls/Allison on the F-35 lift fan clutch.
As for the sprag clutch, they do not like to operate under overrun for extended periods, such as this Sikorsky design concept would require.
Since they have two engines with an interconnect, they would be better off doing a fully synchronized shift using dog clutches with the speeds controlled by the FADEC's.
What model S76?
-- IFMU
-- IFMU
Join Date: Nov 2004
Location: Cambridgeshire, UK
Posts: 1,334
Likes: 0
Received 0 Likes
on
0 Posts
If sightings are X2 then has the new gearbox been installed and ground tested yet? 
----
AATD Drivetrain RFP
The proposal asks for a 55% increase in the drivetrain hosepower to weight ratio compared to US Army year 2000 levels. Any of the drivetrain engineers out there think this is a practical goal?
John, i've given this a little thought, although am still to attend the Cranfield (UK) gear design course (which could make me think again).
The only way the set target is remotely possible is if an epicyclic final reduction was put in just below the rotorhead. Have a look at the differential bevel reduction geartrains driving some propeller installations. With the sort of reduction required for typical helicopter Nr it would likely end up being at least one stage of in plane epicyclic gearsets (with each stage input to sun and output from panet carriers). The potential weight saving would come from having lower torque at the geartrain from each engine (and first reduction stage if two stage reduction was selected).
In terms of materials i would have thought that helicopters have already carefully selected steels with the best tooth contact fatigue life. I'm not very familiar with helicopter gearbox designs - do they already use heat fitted toothed rings over aluminium/titanium hubs?
----
Originally Posted by John Dixson
AATD Drivetrain RFP
The proposal asks for a 55% increase in the drivetrain hosepower to weight ratio compared to US Army year 2000 levels. Any of the drivetrain engineers out there think this is a practical goal?
The only way the set target is remotely possible is if an epicyclic final reduction was put in just below the rotorhead. Have a look at the differential bevel reduction geartrains driving some propeller installations. With the sort of reduction required for typical helicopter Nr it would likely end up being at least one stage of in plane epicyclic gearsets (with each stage input to sun and output from panet carriers). The potential weight saving would come from having lower torque at the geartrain from each engine (and first reduction stage if two stage reduction was selected).
In terms of materials i would have thought that helicopters have already carefully selected steels with the best tooth contact fatigue life. I'm not very familiar with helicopter gearbox designs - do they already use heat fitted toothed rings over aluminium/titanium hubs?
Drivetrain Weight Reduction
Graviman, I should have realized that with an RFP out there, the gearbox/drivetrain designers are hardly likely to respond to questions now.
My skepticism re the 55% number is based on watching what the SA engineers, and those at Eurocopter, Boeing and Bell have been doing with each new model. The improvements have been incremental, both in materials and processes. Durability, reliability and ballistic survivability usually countervail the weight reduction process, and it seems like a 55% number is unachievable with the technology available to the designers. Maybe that is a challenging enough statement to draw some rejoinder!
For certain: someone comes up with a 55% lighter drivetrain on a larger helicopter, one that can take ballistic hits, mis-procedures by the maintenance folks, tolerant to production machining realities, and boast high MTBO's, and you'll be looking at the next Engineering VP at that company!
Thanks,
John Dixson
My skepticism re the 55% number is based on watching what the SA engineers, and those at Eurocopter, Boeing and Bell have been doing with each new model. The improvements have been incremental, both in materials and processes. Durability, reliability and ballistic survivability usually countervail the weight reduction process, and it seems like a 55% number is unachievable with the technology available to the designers. Maybe that is a challenging enough statement to draw some rejoinder!
For certain: someone comes up with a 55% lighter drivetrain on a larger helicopter, one that can take ballistic hits, mis-procedures by the maintenance folks, tolerant to production machining realities, and boast high MTBO's, and you'll be looking at the next Engineering VP at that company!
Thanks,
John Dixson
Join Date: Sep 2007
Location: USA
Posts: 601
Likes: 0
Received 0 Likes
on
0 Posts
MRGB weight reduction
John,
A 55% improvement in the average year 2000 gearbox power to weight metric is definitely achievable with currently available aerospace design tools, materials and manufacturing processes. As far as gear and bearing materials, the best option is X-53 for gears and M50NiL for bearings if you need to meet a 30 minute loss of lube requirement. The serious weight reduction will come from novel and clever arrangements of gears, and not from new materials.
The most effective approach to reducing gearbox weight is by minimizing the amount of gear and bearing steel used. Using a gear configuration that keeps the shaft speeds as high as possible going into the final gear stage is beneficial. Higher shaft speeds means less transmitted torque, and less transmitted torque means smaller gears and bearings. However, this concept requires that your final gear stage has a very high reduction ratio, which is difficult with a conventional planetary.
Another current approach is to employ torque splitting. Dividing the torque among many small driving pinions symmetrically arranged about a large bull gear to give balanced mesh forces, usually ends up being lighter than a single larger mesh with fewer parts. Once again, the difficulty here is getting proper load sharing among the many gears.
The other goals in that AATD program will be harder to meet. Improvements in noise, reliability and support costs can be achieved. But I can't foresee a 35% reduction in manufacturing costs. In fact, I would predict just the opposite occurring. Higher performance and reliability will cost more money, not less.
The X2's MRGB is a good example of where rotorcraft gearboxes are headed. It only has a single engine input, but it has coaxial, counter rotating outputs and multiple speed ranges. MRGB's are becoming more complex, which naturally means they will become more expensive.
Regards,
riff_raff
A 55% improvement in the average year 2000 gearbox power to weight metric is definitely achievable with currently available aerospace design tools, materials and manufacturing processes. As far as gear and bearing materials, the best option is X-53 for gears and M50NiL for bearings if you need to meet a 30 minute loss of lube requirement. The serious weight reduction will come from novel and clever arrangements of gears, and not from new materials.
The most effective approach to reducing gearbox weight is by minimizing the amount of gear and bearing steel used. Using a gear configuration that keeps the shaft speeds as high as possible going into the final gear stage is beneficial. Higher shaft speeds means less transmitted torque, and less transmitted torque means smaller gears and bearings. However, this concept requires that your final gear stage has a very high reduction ratio, which is difficult with a conventional planetary.
Another current approach is to employ torque splitting. Dividing the torque among many small driving pinions symmetrically arranged about a large bull gear to give balanced mesh forces, usually ends up being lighter than a single larger mesh with fewer parts. Once again, the difficulty here is getting proper load sharing among the many gears.
The other goals in that AATD program will be harder to meet. Improvements in noise, reliability and support costs can be achieved. But I can't foresee a 35% reduction in manufacturing costs. In fact, I would predict just the opposite occurring. Higher performance and reliability will cost more money, not less.
The X2's MRGB is a good example of where rotorcraft gearboxes are headed. It only has a single engine input, but it has coaxial, counter rotating outputs and multiple speed ranges. MRGB's are becoming more complex, which naturally means they will become more expensive.
Regards,
riff_raff